Back to 2022 Posters


Patient-specific Simulation Of Mitral Valve Surgical Repair Using An Integrated Left Ventricle Model
Hao Liu1, Alison Pouch2, Joseph Gorman2, Robert Gorman2, Michael Sacks1.
1University of Texas at Austin, Austin, TX, USA, 2The University of Pennsylvania, Philadelphia, PA, USA.

OBJECTIVE: In this study, our focus is to develop patient-specific left-ventricle (LV) - mitral valve (MV) in-silico models for mitral valve surgical repair predictions. The MV surgical repair with undersized ring is currently the preferred treatment for ischemic mitral regurgitation (IMR). However, over 30% of patients treated this way develop significant recurrent IMR within 12 months, a direct consequence of adverse LV remodeling after MI. The primary objective will be to create a simulation pipeline to predict and optimize annuloplasty treatment for individual patients using pre-surgical and population average data.
METHODS: Clinical, real-time three-dimensional echocardiography (rt-3DE) images of patients' mitral valves were segmented by radiologists using a neural network (NN)-based approach to produce high-fidelity LV-MV structure (Figure 1A). Raw data was collected including pressure, cavity volume, and disease information while external population-averaged data including fiber orientation, mechanical properties of soft tissue, and structure of chordae insertions was introduced into patient-specific model. Finally, personalized clinical outcomes and predictions were generalized by finite element simulation (Figure 1B).
RESULTS: Preliminary results of the response of the MV to myocardial infarction (MI) demonstrated an increase of annular dilation and leaflet tethering. Furthermore, loss of active contraction due to different MI locations plays a key role in MV regurgitation.
CONCLUSIONS: We present a pipeline that allows for patient-specific modeling of lv-mv to predict the valvular response to annuloplasty repair and with high predictive power and onset of ischemic regurgitation followed by myocardial infarction. Our framework only relies on the clinically obtainable imaging data prior to the MV repair operation and thus can be extended into a virtual surgery tool that provides surgeons with additional insight into the patient-specific valvular response to different treatment options.


Back to 2022 Posters